Cleaner heavy transports : Environmental and economic analysis of liquefied natural gas and biomethane

Cleaner heavy transports e Environmental and economic analysis of

lique

fied natural gas and biomethane

Marcus Gustafsson

, Niclas Svensson

Environmental Technology and Management, Department of Management and Engineering, Link€oping University, SE-581 83, Link€oping, Sweden

a r t i c l e i n f o

Received 21 April 2020 Received in revised form 2 July 2020

Accepted 29 July 2020 Available online 9 August 2020

Handling editor: Yutao Wang

Keywords: Biomethane Natural gas Heavy transport Liquefaction Life cycle assessment Life cycle cost

a b s t r a c t

Looking to reduce climate change impact and particle emissions, the heavy-duty transport sector is moving towards a growth within technology and infrastructure for use of liquefied natural gas (LNG). This opens an opportunity for the biogas market to grow as well, especially in the form of liquefied biomethane (LBM). However, there is a need to investigate the economic conditions and the possible environmental benefits of using LBM rather than LNG or diesel in heavy transports. This study presents a comparison of well-to-wheel scenarios for production, distribution and use of LBM, LNG and diesel, assessing both environmental and economic aspects in a life cycle perspective. The results show that while LNG can increase the climate change impact compared to diesel by up to 10%, LBM can greatly reduce the environmental impact compared to both LNG and diesel. With a German electricity mix, the climate change impact can be reduced by 45e70% compared to diesel with LBM from manure, and by 50 e75% with LBM from food waste. If digestate is used to replace mineral fertilizer, the impact of LBM can even be less than 0. However, the results vary a lot depending on the type of feedstock, the electricity system and whether the calculations are done according to RED or ISO guidelines. Economically, it can be hard for LBM to compete with LNG, due to relatively high production costs, and some form of economic incentives are likely required.

© 2020 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

1. Introduction

Transportation of people and goods requires vast amounts of energy and is a major source of air pollution and greenhouse gas (GHG) emissions. Within the European Union (EU), there is a goal to reduce GHG emissions from transports by at least 60% by 2050

compared to 1990, without impairing mobility (European

Commission, 2016). Around 25% of the GHG emissions from road transports in the EU are accounted for by commercial heavy-duty

vehicles such as buses and trucks (EEA, 2018). Alternative

propul-sion technologies to reduce pollution and climate change impact are broadly investigated in research, including batteries, fuel cells

and gas. When it comes to gas, liquefied natural gas (LNG) has the

advantage of a higher volumetric energy density than compressed

gas (Benjaminsson and Nilsson, 2009), which makes it feasible to

use in heavy and long-distance transports. Engines designed for methane propulsion are currently being produced by several truck

and bus manufacturers, including Daimler (2018), Isuzu (2018),

Iveco (2015),Scania (2017)andVolvo Trucks (2017). At the same time, efforts to build up an LNG fueling infrastructure take place e.g.

within the LNG Blue Corridors project (LNG BC, 2013) and through

current fueling station expansion in the Nordic countries (Gasum,

2019).Smajla et al. (2019)stated that LNG in heavy transports is more economic than diesel in a long perspective and that switching to LNG is crucial for meeting environmental goals.

Compared to diesel, compressed or liquefied natural gas is often

shown to give a certain reduction of GHG emissions from vehicles. Considering the whole production-to-use pathway of

fuelsdu-sually referred to as well-to-wheel (WTW) analysisdSpeirs et al.

(2020) showed a reduction of GHG emissions with natural gas

compared to diesel of up to 16%,Arteconi et al. (2010)reported a

GHG emission reduction with LNG of up to 10%,B€orjesson et al.

(2016)showed a reduction of 2e12% for heavy-duty vehicles and

Ou and Zhang (2013)found a GHG reduction of 5e10%. However,

the GHG savings are limited by a lower efficiency of gas engines

compared to diesel engines, by methane leakage and by the fact

that natural gas is a fossil fuel.Arteconi et al. (2010)found that

small-scale LNG production resulted in GHG emissions in line with

* Corresponding author.

E-mail addresses:marcus.gustafsson@liu.se (M. Gustafsson),niclas.svensson@ liu.se(N. Svensson).

Contents lists available atScienceDirect

Journal of Cleaner Production

j o u r n a l h o me p a g e :w w w .e l se v i e r. co m/ lo ca t e / jc le p r o

https://doi.org/10.1016/j.jclepro.2020.123535

diesel, andStettler et al. (2019)found that the GHG emissions of heavy-duty vehicles running on natural gas were equal to or up to

8% higher than the emissions of equivalent diesel vehicles.Speirs

et al. (2020)also found that the GHG emissions could in fact

in-crease by replacing diesel with natural gas, andCooper et al. (2019)

found that the benefits of reduced climate change impact with

natural gas could be lost if the methane emissions exceed 1.5e3.5%. From a technical point of view, fossil natural gas can easily be exchanged for renewable biomethane, either as compressed or

liquefied biomethane (LBM, also known as bio-LNG). Biomethane is

produced from organic materials through fermentation (biogas) or thermal processes and is treated to have a similar properties as natural gas, and can therefore be regarded as one possible way

towards a more renewable gas network (Speirs et al., 2018). While

the EU market for natural gas is largely dependent on imports (Eurostat, 2019a; IGU, 2017), biogas and biomethane are often

produced and used domestically or even locally (Kampman et al.,

2016). Currently, the practice of using biomethane in road

trans-ports is very limited, with Sweden being the leader in Europe (EurObserv’ER, 2017), although there is a growing interest and

policy support in many countries (Gustafsson and Anderberg,

2020). On average, the share of biomethane in European vehicle

gas is about 17%, with some countries reaching close to 100% (H€ormann, 2020). Estimates suggest a potential steep increase in

LBM production in Europe (Agelbratt and Berggren, 2015).

P€a€akk€onen et al. (2019)concluded that half the heavy-duty trans-port sector in Finland could rely on biomethane by 2030. The total biogas production in the EU-28 has increased by a factor 10 since the turn of the millennium, from around 70 PJ/year to over 700 PJ/

year (Eurostat, 2019b). While the biogas production from landfill

gas and sewage sludge have been rather static in the last 15 years, most of this change is due to a growth in other anaerobic digestion,

such as energy crops, manure and food waste (Eurostat, 2019b). For

the further development of biofuel production, the European Union

advocates second-generation, or advanced, biofuels (European

Commission, 2009), which are based on non-food crops or

byproducts from agriculture and forestry (Sims et al., 2008).

Second-generation biofuels should thereby achieve significant

reductions of environmental impact compared to fossil fuels, while also avoiding a competition of land between food and fuel pro-duction. While large producers of biogas like Germany and Italy

currently use a lot of energy crops in their supply chains (EBA,

2020), many EU countries even exclude energy crops in national

regulations on biogas production (EurObserv’ER, 2017).

The environmental impact of biomethane as a transport fuel has

been assessed in several studies. Shanmugam et al. (2018)

compared the environmental impact of LBM and diesel for heavy trucks in Sweden, concluding that LBM is superior in 7 out of 10

investigated impact categories. Hagos and Ahlgren (2018)

compared the energy balance and GHG emissions of natural gas and biomethane in road and maritime transports, showing that biomethane can greatly reduce the WTW emissions despite a

higher energy input.Lyng and Brekke (2019)performed a

well-to-wheel analysis of fuels for buses and found that biomethane from food waste or manure is among the vehicle fuels with the lowest

environmental impacts on the market. Further,Natividad

Perez-Camacho et al. (2019)found that biomethane can reduce the GHG

emissions by around 500 kg CO₂-equivalents per MWh in a life cycle

perspective compared to petrol or diesel. In addition to the envi-ronmental aspects of biomethane as a vehicle fuel, biogas and biomethane production systems typically include environmental

and other benefits that are not strictly tied to the production of an

energy carrier (Hagman and Eklund, 2016). For example,

produc-tion of biogas through anaerobic digesproduc-tion generates a by-product that can be used as a fertilizer, thus displacing an equivalent amount of conventional mineral fertilizer. The technical reports by

Edwards et al. (2014)andB€orjesson et al. (2016)both note that not only is the possible GHG reduction compared to diesel substantially larger with biomethane than with natural gas, but the climate change impact of biomethane can even become negative if external effects of biogas production are included in the calculations. While this type of system expansion is encouraged in the ISO standards

for LCA (ISO, 2006a,2006b), the EU Renewable Energy Directive

(RED) does not allow it (European Commission, 2009).

Thefinancial conditions for production of biomethane can vary a

lot depending on technology, scale, feedstock, energy costs, Nomenclature

Abbreviations

AD anaerobic digestion

AF annuity factor

AS amine scrubbing

LBM liquefied biomethane

C3MR propane/mixed refrigerant

GHG greenhouse gas

LCA life cycle assessment

LCC life cycle cost

LNG liquefied natural gas

MCF methane conversion factor

MR mixed refrigerant

PM particulate matters

RED renewable energy directive

SI spark ignition

TTW tank-to-wheel

VOC volatile organic compounds

VS volatile solids

WS water scrubbing

WTW well-to-wheel

Symbols

B0 maximum methane potential

CH₄ methane

CO carbon monoxide

CO₂ carbon dioxide

i interest rate

mCH4 mass of methane emitted during storage

mmanure mass of manure

N depreciation period

N₂ nitrogen

NH_₃ ammonia

NH₄-N ammonium nitrogen

N₂O dinitrogen monoxide

NOₓ nitrogen oxides

SO₂ sulfur oxides

rCH4

s

economic incentives and other factors. Several studies have assessed the economic aspects of different ways of producing

biogas and biomethane, including recent publications byLombardi

and Francini (2020),Martín-Hernandez et al. (2020)andGustafsson et al. (2020a). However, comparisons against competing technol-ogies such as fossil fuels are less common in the literature.

P€a€akk€onen et al. (2019)found that in Finland, biomethane pro-duction costs could be competitive with the market price of diesel, not taking into account any possible margin for revenue from

selling the biomethane.B€orjesson et al. (2016)found that the WTW

costs of heavy-duty trucks with LBM could be comparable to diesel trucks in Sweden, including taxes, but not as low as with LNG.

Tybirk et al. (2018)argued that it will be difficult for biomethane producers to compete with the price of natural gas, especially in regions with availability of shale gas. Still, there is a need for further studies on how factors such as technology, feedstock and scale

influence the competitiveness of biomethane as a vehicle fuel.

Liquefaction of methane is a well-known concept with proven technologies. However, they are usually designed for the scale of natural gas production. Biogas is produced in much smaller vol-umes, and in comparison to LNG production facilities a liquefaction

plant for biomethane could be considered “nano-scale” (Tybirk

et al., 2018). Natural gas liquefaction processes worldwide are

mainly based on Air Products’ technologies with mixed refrigerant

condensation (IGU, 2017). The most common liquefaction

tech-nology for natural gas applications is C3MR, which utilizes propane

(C_{₃H₈) and a mixed refrigerant (MR) for cooling and condensing the}

gas in two cycles (Usama et al., 2011). Gustafsson et al. (2020a)

analyzed the costs, energy use and environmental impact of

sce-narios for production and distribution of compressed and liquefied

biomethane, including thefive upgrading technologies described

byBauer et al. (2013)and four different technologies for liquefac-tion. Their results showed that the differences between the com-mercial upgrading technologies are quite small, except for amine scrubbing which uses less electricity and more heat than other

technologies (Gustafsson et al., 2020a). In general, biomethane

production systems are more dependent on electricity than natural

gas production (Edwards et al., 2014), where part of the gas stream

is often used in gas turbines for internal energy demands (Raj et al.,

2016;Songhurst, 2018). This implies that the local electricity

sys-tem would influence the WTW performance of biomethane

compared to natural gas or diesel, which (to the authors’

knowl-edge) is yet to be assessed in research.

While there has been some previous research on the environ-mental aspects of LNG and LBM in transports, there is a lack of studies comparing the economic aspect. Moreover, there appears to be a lack of attention towards geographical differences in

com-parisons of LNG and LBM, specifically the influence on the

envi-ronmental performance of the local electricity system. This paper presents an environmental and economic WTW analysis of sce-narios for production and distribution of LNG and LBM for heavy-duty trucks in a European context. The comparison includes different feedstock for biogas production as well as different technologies to produce and distribute LNG and LBM, and different electricity mixes for production of LBM. As a reference, conven-tional diesel is also included in the comparison. Environmental

impact is assessed in form of climate change, acidification,

eutro-phication, ground-level ozone formation and stratospheric ozone layer depletion, both with the RED method and with system expansion according to ISO. The cost of producing LBM is calculated depending on production capacity and compared to the market price of LNG. Thus, this paper provides a broad overview of the environmental and economic aspects of LNG and LBM for heavy-duty trucks, covering several possible pathways and assessing the

influential factors.

2. Methodology

The analysis consisted of two parts, as illustrated inFig. 1: an

analysis of WTW environmental impact of liquefied biomethane

(LBM), liquefied natural gas (LNG) and diesel, which was done

through life cycle assessment (LCA); and a comparison between the WTW life cycle cost of LBM production and the market price of LNG. The WTW analysis was done by modelling and analyzing a set of scenarios of production, distribution and use. Each scenario rep-resents a possible pathway for the fuel from resource extraction to use in the vehicle, including all the intermediary processes to get there. The modelling process and the studied scenarios are

pre-sented in sections2.1 and 2.2, respectively. The data used in

envi-ronmental and economic calculations is presented in section2.3,

and the sensitivity and uncertainty analyses conducted are further

explained in section2.4.

2.1. Modelling

The environmental impact of the scenarios was evaluated

following ISO 14040 and 14,044 guidelines for LCA (ISO, 2006a,

2006b). An LCA consists of four principal steps: Goal and scope

definition, Life cycle inventory (data collection), Environmental

impact assessment and Interpretation of results. The boundaries were set to cover the whole fuel pathway from extraction or

collection of raw materials to combustion in the vehicle’s engine, in

other words well-to-wheel (WTW). Modelling and calculations

were done using SimaPro 8, a widely used tool for LCA (Goedkoop

et al., 2016). Environmental impact assessment, i.e. calculation of

environmental impact of the included materials and energyflows,

was done with the ReCiPe heuristic midpoint method (PRe, 2014),

which is commonly used in LCA research studies. The assessment

coveredfive impact categories: climate change, terrestrial

acidifi-cation, freshwater eutrophiacidifi-cation, ozone formation (impact on

Fig. 1. Overview of the methodology of the study. The analysis consisted of environ-mental impact assessment and economic analysis.

human health) and stratospheric ozone depletion. For processes

where specific data could not be obtaineddsuch as production of

diesel and natural gas, fuel distribution and electricity pro-ductiondgeneric processes from the Ecoinvent database version

3.5 (Moreno Ruiz et al., 2018;Weidema et al., 2013) were used.

These processes were selected to represent the modelled system as

closely as possible, e.g. electricity markets in specific countries (see

section2.4) and distribution trucks of European standard (Euro 6).

In the life cycle cost (LCC) calculations, costs of capital, operation and maintenance and distribution for LBM pathways were compared to the European market price of LNG, as the equivalent costs for LNG pathways could not be retrieved. This means that the LBM would have to add a revenue for sales to be able to compare with LNG on the same basis. However, in this study the comparison is used to see how close LBM is to becoming a viable market option in terms of price and thus the difference can be used for discussion. The LBM costs were calculated using the annuity factor (AF), with a depreciation period (N) of 15 years and a 6% interest rate (i):

AF¼ i

1 ð1  iÞN (1)

Apart from cost for energy, operation and maintenance costs were assumed to be 2.5% of the investment cost. Similar economic assumptions were found in several comparable studies, including

Larsson et al. (2015), B€orjesson et al. (2016), Gustafsson et al. (2020a). The cost of distribution of LBM by truck was set to 1.80 V/km, including diesel, driver, truck and other related costs, plus

64.50V/hour for loading and unloading (4 h per delivery).

Distri-bution costs were based onPettersson et al. (2006)and updated for

inflation. The cost calculations for fueling LBM did not include costs

for building and operating the fueling station, only electricity for the pump, since the same fueling stations are also used for LNG, and the volumes of LNG are much higher compared to LBM. The same applies for distribution by gas grid.

2.2. Studied scenarios

The studied WTW scenarios covered the way from raw material to use in a vehicle in a European road transport system. The envi-ronmental impact of LBM, LNG and diesel were compared on the basis of 1 km transport with a 40-ton long haulage truck, taking into consideration differences in fuel economy of gas and diesel engines. In the economic comparison between LBM and diesel, the cost of producing 1 kWh LBM was compared to the market price of 1 kWh LNG.

For LBM, the system boundary started at collection of raw ma-terial for anaerobic digestion (AD) and ended at combustion in the

vehicle (Fig. 2), in a spark-ignited (SI) Otto engine. Two types of

feedstock were considered for the biogas production: manure and food waste, which are both eligible feedstock for advanced or second-generation biofuels. The biogas from the AD was assumed to be upgraded to biomethane through water scrubbing (WS) or

amine scrubbing (AS) after which it was liquefied through

mixed-refrigerant (MR) or nitrogen (N₂) cycle, distributed by truck,

fueled and combusted in a vehicle. Alternatively, the biomethane was distributed via a high-pressure gas grid to a pressure reduction liquefaction facility. WS is currently the most common technology

for biogas upgrading in Europe in terms of produced volume (Prussi

et al., 2019), while AS stands out compared to other conventional

upgrading technologies when it comes to energy use (Gustafsson

et al., 2020a). For gas upgraded through WS, an extra CO₂

polish-ing step was considered before MR or N₂ cycle liquefaction whereas

AS can meet the purity requirements for liquefaction at the cost of

an increased energy use (Bauer et al., 2013; Karlsson, 2018).

Distribution of lique_{fied gas was done by truck with a 25-ton}

cryogenic tank. For scenarios with pressure reduction liquefac-tion, the gas was compressed to 60 bar and distributed via the natural gas grid. The pressure reduction liquefaction process was

assumed to produce 10% liquefied gas, while the remaining 90%

was considered to be used as in gaseous form in other applications (He and Ju, 2013).

For the LBM scenarios, two sets of calculations were done, using different system boundaries: one following the ISO standard for LCA, and one following RED. In the ISO calculations, the use of digestate from anaerobic digestion (AD) as a biofertilizer were included, and alternative handling of the substrates used in AD was considered. In the case of manure as substrate, this meant that methane emissions from conventional manure handling could be avoided, and a larger amount of mineral fertilizers could be replaced as the AD process increases the nitrogen availability in the digestate. In the food waste scenarios, the alternative waste handling scenario was assumed to be incineration with cogenera-tion of heat and electricity, and the digestate from AD was also in this case used as fertilizer. In the RED calculations, the environ-mental impact of the digestate, as well as alternative waste handling and the possible gain of displacing mineral fertilizer production, was not included, as RED prescribes energy allocation between biogas and digestate rather than system expansion.

For LNG, the system boundary covered the way from raw

ma-terial extraction to combustion in the SI engine (Fig. 3), with similar

pathways as for LBM. Before liquefaction, natural gas goes through

a series of purification processes to clean it from carbon dioxide,

hydrogen sulfide and water (Ma et al., 2017). The studied

lique-faction technologies include C3MR and N₂ cycle liquefaction, as well

as pressure reduction liquefaction from the gas grid.

In the food waste LBM scenarios, substrate and digestate were assumed to be transported on average 25 km between the substrate source and the biogas plant, and between the biogas plant and the digestate recipient. An exception was made in the manure sce-narios, where the biogas plant was assumed to be located at the farm, making such transports redundant. The mineral fertilizer replaced in ISO calculations was assumed to be produced further from the area of use, requiring a transport of 50 km. LNG and LBM were transported in a 25-ton cryogenic tank over an average dis-tance of 200 km. The LNG pathways additionally included a ship transport over 10,000 km, representing the distance between Middle East (Qatar) and Western Europe (Netherlands) via the Suez Canal. Alternatively, the fuels were distributed via pipeline, over a distance of 200 km for biomethane and 5000 km for natural gas. Diesel was assumed to be produced in the Middle East and

trans-ported the same way as LNG from the refinery to the area of use

(Fig. 4).

2.3. Data collection

Energy use, costs, methane content in biogas and methane slip

for anaerobic digestion are listed inTable 1, and data for substrates

and digestates is listed inTable 2. Data for manure were taken to be

the average of data for cattle and swine manure. The share of

ammonium nitrogen (NH_{₄-N) was assumed to be 20% higher after}

anaerobic digestion compared to the undigested substrate. This

value falls within intervals found in publications byM€oller and

Müller (2012)and Sinclair et al. (2013). In accordance with the same sources, the availability of phosphorus and potassium was assumed to not increase through anaerobic digestion. In ISO cal-culations, digestate of manure was therefore assumed to replace an

equal amount (mass input) of undigested manure, plus artificial

nitrogen fertilizer corresponding to the increased NH₄-N content,

be unchanged compared to the alternative waste handling and were left out of the calculations. In the food waste scenarios, the digestate was assumed to replace mineral fertilizer corresponding

to the total NH₄-N content of the digestate.

Spreading of the digestate was assumed to cause 10% higher

emissions of ammonia (NH₃) and dinitrogen monoxide (N₂O)

compared to mineral fertilizers (Tufvesson et al., 2013). The NH₃

emissions to air were set to 5% of the total ammonium content of

the digestate (Lantz et al., 2009), and 1% was assumed to be emitted

in the form of N₂O (De Klein et al., 2006). The remaining ammonium

in the digestate was modelled as emission to soil. Methane emis-sions from digestate spread on agricultural land were assumed to be negligible in relation to methane emissions from biogas

pro-duction (Amon et al., 2006;Tufvesson et al., 2013). The share of

biogasflared in the AD plant due to occasional mismatch between

biogas production and demand was set to 5%. Internal heat demand

Fig. 2. Schematic view of the modelled WTW scenarios for LBM. Going from the left to the right, each of the indicated pathways represent a possible scenario from production to end user: biogas production, upgrading to biomethane, liquefaction and distribution, fueling and combustion in vehicle engine. The color and weight of the lines indicate the state (gas/liquid) and methane content at that part of the pathway. (For interpretation of the references to color in thisfigure legend, the reader is referred to the Web version of this article.)

Fig. 3. Schematic view of the modelled WTW scenarios for LNG. Going from the left to the right, each of the indicated pathways represent a possible scenario from production to end user: gas extraction and processing, liquefaction and distribution, fueling and combustion in vehicle engine. The color and weight of the lines indicate the state (gas/liquid) and methane content at that part of the pathway. (For interpretation of the references to color in thisfigure legend, the reader is referred to the Web version of this article.)

for the AD plant, as well as for AS upgrading, was assumed to be

covered through combustion of raw biogas with 95% efficiency

(Lantz et al., 2009;Tufvesson et al., 2013).

In the scenarios where manure was used as a feedstock, the AD process was assumed to reduce methane emissions compared to

conventional handling of manure by 90% (Tufvesson et al., 2013).

The amount of methane emissions from manure was calculated through:

mCH4¼ mmanure,VS,B0,MCF,

r

where mmanureis the amount of manure (kg), VS is the share of

volatile solids (%), B0is the maximum methane production of the

substrate (m3_{/kg), MCF is the methane conversion factor (%),}

_rCH4

the density of methane (kg/m3_{) at normal temperature and}

pres-sure and mCH4is the amount of methane emitted during storage

(kg). B0was set to 0.345 (average between dairy cattle and swine

manure) in accordance withGavrilova et al. (2006), and MCF to

3.5% (Tufvesson et al., 2013).

In the scenarios where food waste was used as a feedstock, the alternative waste handling was considered to be combustion with

energy recovery, with heat and electricity efficiencies of 74% and

21%, respectively (Hamelin et al., 2014), and a lower heating value

of 20 MJ/kg dry matter (Davidsson et al., 2007). As the food waste in

these scenarios was used for biogas production rather than energy production, equivalent amounts of heat and electricity were considered to be produced from natural gas.

Data for the other processes included in the studied

scenar-iosdgas upgrading, liquefaction and fuelingdare listed inTable 3.

Internal electricity demand for LNG production was assumed to be

covered with gas turbines, using part of the gas flow (Raj et al.,

2016;Songhurst, 2018), with an ef_{ficiency of 40% (}Siemens, 2016). The production of natural gas was assumed to have a methane slip of 1%, similar to water scrubber biogas upgrading. In both cases, the

off-gas was assumed to be treated through catalytic oxidation,

converting 95% of the methane to carbon dioxide and water (Herbst

et al., 2010). The electricity demand for gas sweetening and dehy-dration was considered to be 10% of the electricity demand for

liquefaction (H€onig et al., 2019), which was set to 0.26 kWh/kg LNG.

This value is in line with data reported byYoon et al. (2009)and

CNPC (2019), and only slightly higher than the 0.23 kWh/kg

re-ported byHajji et al. (2019). For LBM production, a German

elec-tricity mix was assumed in the default case, and other alternatives were included in a sensitivity analysis. Road transports were considered to be done with gas vehicles in the gas scenarios (using LNG in the LNG pathways and LBM in the LBM pathways) and with diesel vehicles in the diesel scenario. Tank-to-wheel (TTW)

emis-sions from combustion of the studied fuels (Table 4) were based on

emission data for heavy freight trucks fromStettler et al. (2019).

The SI gas engines were considered to require 18% more energy per

km than the diesel engines (B€orjesson et al., 2016;Gustafsson et al.,

2020b): 3.24 kWh/km, compared to 2.74 kWh/km for diesel (Gustafsson et al., 2020b).

In the economic analysis, the cost for producing LBM was compared to the European market price of LNG, which was found to

have varied between 4 and 10 USD/mmBtu over the lastfive years,

or 0.012 to 0.030 EUR/kWh (YCharts, 2019).

2.4. Sensitivity and uncertainty analysis

Electricity use for production of LBM was modelled with elec-tricity mixes representative for Germany, UK, Italy and Sweden, in order to investigate the sensitivity with respect to how the elec-tricity is produced. Germany, UK and Italy are the three largest biogas producers in Europe, while Sweden currently has the largest

market for biomethane as a transport fuel (EurObserv’ER, 2017). In

the Ecoinvent database version 3.5 (updated 2018), the major components of the studied electricity mixes are: lignite (28%), hard coal (22%) and nuclear power (18%) in the German mix; hard coal (34%), natural gas (28%) and nuclear power (21%) in the UK mix; natural gas (34%), hydropower (26%) and hard coal (18%) in the Italian mix; and hydropower (45%) and nuclear power (43%) in the

Swedish mix (Table 5) (Moreno Ruiz et al., 2018;Weidema et al.,

2013). Considering the fact that electricity is exported and

im-ported between different countries in Europe, these scenarios serve more to illustrate the difference between contrasting electricity systems than to represent the situation in either of the four countries.

For climate change impact, the level of uncertainty in the results was assessed through Monte Carlo analysis, which constructs an interval based on probability functions and the combined

Table 1

Data used in modelling of environmental impact and life cycle costs of anaerobic digestion.

Parameter Value Unit

Electricity 30a _{MJ/ton substrate}

Heat 90a _{MJ/ton substrate}

Methane content 60%b

Methane slip 0.5%c

Investment cost 240e340d _V/m3

reactor volume

a_{(Lantz et al., 2009).} b_{(Rasi et al., 2007).} c _{(Lantz and Bj€ornsson, 2016).} d _{(Lantz, 2013).}

Table 2

Data for substrates and digestates used in modelling of anaerobic digestion. The table shows the used values (average) as well as the Intervals found in literature. Substrate type TS, % VS of TS, % Methane yield, m3

/ton VS Total N, kg/ton NH₄-N in digestate, % of N

Manure 8 80 290 3.9 67

Cattle 8.5e9a,b,c ₈₀a _213e240a,d _2.6e4.3a,b,e,f,g _54e67e,g

Swine 7.3e8a,b ₈₀a _268e450a,d _3.1e5.9a,b,e,f _68e81a,e

Food waste 22 (18e28)h _{90 (87e94)}h _{360 (275e461)}a,h,i _{5.2 (3.1e7.5)}a,e,h,j _{70 (60e80)}e,j

a_{(Carlsson and Uldal, 2009).} b_{(Kirchmann and Witter, 1992).} c _{(Walsh et al., 2012).} d _{(Gavrilova et al., 2006).} e_{(Sinclair et al., 2013).} f _{(M€oller and Müller, 2012).} g_{(Lukehurst et al., 2010).} h_{(Banks et al., 2018).}

i_{(Davidsson et al., 2007).} j_{(Ljung et al., 2012).}

uncertainty of all parameters. Parameter values were varied using a lognormal distribution, where the value of a parameter lies

be-tween the mean value divided by the variance (

s

value multiplied by

s

were considered to be fairly reliable and representative and

s

set to 1.1. A higher value, 1.2, was used for parameters regarding the AD process (electricity and water use, amount of digestate) and for

transports, as they were assumed to vary more. For methane slip,

s

was set to 1.5.

In the economic calculations, a sensitivity analysis was done by varying the utilization of production capacity and the distribution

distance for the LBM pathways. The availability of biogas upgrading plants with respect to maintenance is generally quite high, around

95% (Bauer et al., 2013;Patterson et al., 2011;Persson, 2003), which

was therefore used as one scenario. To account for possible limiting factors (e.g. access to substrate, variations in demand), a lower utilization rate of 50% was also evaluated. For distribution by

semi-trailer, the distance was varied between 0 km (filling station

adja-cent to production plant) and 400 km (double the default value).

3. Results and analysis

3.1. Environmental impact assessment

Fig. 5shows the climate change impact per km of the studied

pathways of diesel, liquefied natural gas (LNG) and liquefied

bio-methane (LBM) with biogas from manure, divided into impact from

different processes and with error bars indicating the 95% con

fi-dence interval of the uncertainty analysis. The best-performing liquefaction technology for LBM is liquefaction by pressure reduc-tion of methane from the high-pressure gas grid. Out of the two off-grid technologies, mixed-refrigerant (MR) liquefaction has a lower energy use and climate change impact than nitrogen cycle lique-faction, both for LNG (C3MR) and for LBM. For biogas upgrading, amine scrubbing gives a slightly lower climate change impact than water scrubbing, due to a lower electricity use.

Compared to diesel, LBM from anaerobic digestion (AD) of manure can reduce the WTW climate change impact of driving a 40-ton long haulage truck by 45e70%, calculated according to the Renewable Energy Directive (RED). If the digestate from AD is

Table 3

Data used in modelling of environmental impact and life cycle costs of pathways for biomethane and natural gas.

Process Electricity Heat Water Chemicals Methane content Methane slip Investment cost

MJ/MJ gas MJ/MJ gas m3_{/kg gas kg/kg gas % % V/(Nm}3_/h)

Water scrubber 0.039a _4.5E-04a _6.0E-05a _98%a _1%a _1,490e5,820a

Amine scrubber 1,630e3,300a

Gas grid quality 0.023a,b _0.017b _6.1E-05a _6.1E-05a _99.8%a,b _0.06%a,b

Liquefaction quality 0.034b _0.017b _6.1E-05a _6.1E-05a _99.995%a,b _0.06%a,b

Catalytic oxidation of off-gas 0.008a

BG cleaning for liquefaction 0.015b _99.995%b _1,400e2,600a

NG cleaning for liquefaction 0.002c

Mixed-refrigerant cycle liquefaction (BG) 0.058b,d _99.995%b _0%e _4,650e21,000f

Nitrogen cycle liquefaction (BG) 0.072b _99.995%b _0%e _8,025a

C3MR cycle liquefaction (NG) 0.019g,h _99.995%b _0%e

Nitrogen cycle liquefaction (NG) 0.043i _99.995%b _0%e

Pressure reduction liquefaction 0.001j

Fueling, LBM/LNG 0.0003k a_{(Bauer et al., 2013).} b _{(Karlsson, 2018).} c _{(H€onig et al., 2019).} d _{(Cryo Pur, 2019).} e _{(Tybirk et al., 2018).}

f _{(Olgemar and Partoft, 2017).} g _{(Yoon et al., 2009).} h_{(CNPC, 2019).} i _(Cryostar). j _{(Patterson et al., 2011).} k _{(Heisch, 2012).} Table 4

Tank-to-wheel emission factors for a heavy-duty truck running on diesel or CNG (SI engine). Based on data fromStettler et al. (2019). TTW emissions, kg/kg fuel

CO₂ CO CH₄ N₂O PM VOC NOx

Diesel 6.87E-02 8.16E-05 e 1.69E-06 2.71E-07 1.81E-07 1.35E-04

LNG (SI) 5.85E-02 1.90E-04 5.04E-05 2.04E-07 1.87E-07 1.58E-06 1.50E-05

Table 5

Energy sources in electricity production mixes in Germany, UK, Italy and Sweden. Based on data from Ecoinvent 3.5 (Moreno Ruiz et al., 2018;Weidema et al., 2013).

Share of production mix, %

Energy source Germany UK Italy Sweden

Hard coal 22 34 18 0.3 Hydropower 5 3 26 45 Natural gas 7 28 34 0.2 Nuclear power 18 21 0 43 Oil 0.3 0.2 5 0.1 Peat 0 0 0 0.1 Wind 11 9 7 8 Biogas 6 0 5 0 Wood chips 2 5 2 3

Blast furnace gas 0.4 0 1 0.2

Coal gas 0.1 0 0.3 0

Geothermal 0 0 2 0

assumed to replace mineral fertilizer in agricultural applications (in line with ISO standard for life cycle assessment), the climate change impact even becomes negative, equivalent to a reduction of 100e125% compared to diesel. For LNG, the WTW climate change impact is comparable to or slightly higher (5e10%) than diesel. This

is mainly due to the higher efficiency of the diesel engine, resulting

in a higher energy demand in the LNG and LBM scenarios (18% higher per km).

The processes contributing the most to the climate change impact of the LBM scenarios, apart from the possible negative impact of digestate, are electricity for AD, upgrading and

liquefac-tion and methane leakages. The uncertainty intervals are±25e35%

for the LNG scenarios, and around ±40e80% for LBM scenarios,

much due to the high variance of methane leakages (

s

also the variance of factors related to substrate and digestate

handling (

s

The climate change impact of the pathways of LBM from food

waste, as well as diesel and LNG, is shown inFig. 6. Just as with

manure-based biogas, pipeline distribution and pressure reduction liquefaction gives the lowest climate change impact for LBM, and

MR results in lower impact than N_{₂ for both LNG and LBM.}

Using a substrate with a higher methane yield, such as food waste, the WTW climate change impact reduction compared to diesel can be as high as 50e75% with RED calculations. With ISO calculations, the reduction compared to diesel is 80e105%. The

uncertainty intervals are±15e40% for the LBM scenarios. The lower

potential methane leakage compared to the manure-based path-ways greatly reduces the uncertainty.

Fig. 7 shows the impact on climate change, acidification, eutrophication, ground-level ozone formation and stratospheric ozone depletion of diesel, LNG and LBM pathways for different electricity mixes. The LNG and LBM pathways shown are the ones that were found to have highest and lowest environmental impact in all categories: C3MR liquefaction and pressure reduction for LNG,

and WSþ N₂ and AS þ pressure reduction for LBM, respectively.

Among the LBM scenarios, ASþ pressure reduction had the lowest

impact and WSþ N₂ had the highest throughout all of the included

impact categories. For the LNG scenarios, it varied which one had the highest and which one had the lowest impact, but the

differ-ences were quite small and C3MR was always slightly lower than N₂

liquefaction.

Overall, the Swedish electricity mix results in the lowest envi-ronmental impact for the LBM scenarios, followed by Germany, Italy and UK. Another general observation is that the ISO method gives lower results than the RED method, except in terms of

acid-ification, where the situation is the opposite, and eutrophication,

where there is almost no difference between ISO and RED. The largest difference is seen in the climate change and stratospheric ozone depletion impact categories.

The climate change impact is lower for all the LBM pathways than for LNG and diesel. The lowest impact is achieved with a Swedish electricity production mix, followed by Italy, UK and

Germany in close competition. As previously seen inFigs. 5and6,

electricity use accounts for a large share of the climate change impact of the LBM pathways, and the use of digestate as fertilizer can greatly reduce the impact. Thus, most scenarios with LBM from manure have a negative climate change impact with ISO

calcula-tion, except for WS þ N₂ with a German (manure, food waste),

Italian (food waste) or UK (food waste) electricity production system.

The potential acidification impact of the studied LBM scenarios

is comparable to LNG and diesel if calculations are done according to RED, and higher if the ISO method is employed. The difference between the four electricity production mixes is smaller than in the

climate change category. The main contributing process to acidi

fi-cation are electricity use and transports. With system expansion according to ISO, the avoided production of mineral fertilizer compensates for some of these factors, but this effect is outranked

Fig. 5. Climate change impact of diesel, LNG and LBM well-to-wheel pathways. The LBM pathways are based on biogas from anaerobic digestion of manure and electricity use is modelled with a German electricity mix.

by ammonia emissions from the biofertilizer. The LNG scenarios

have a lower acidification impact than diesel.

The eutrophication impact is higher than LNG and diesel for most of the LBM pathways with a German or UK electricity pro-duction system, but comparable or lower with Swedish or Italian electricity. The LNG scenarios have a higher or equal impact to diesel. The German electricity production system has a much higher eutrophication impact per kWh electricity than the others (a factor 4, 6 and 112 compared to UK, Italy and Sweden, respectively). Apart from the electricity, the processes contributing the most to eutrophication in the LBM scenarios are the transports. With sys-tem boundaries according to ISO, the use of biofertilizer to replace mineral fertilizer contributes with a small reduction of the eutro-phication impact, whichdwith a Swedish electricity mixdis enough to result in a negative impact.

Looking at ground-level ozone formation, the impact is lower for the LBM scenarios than for LNG and diesel, except for a few cases with a UK, Italian or German electricity production system. LNG has a higher impact than diesel with pressure reduction liquefaction, but lower than diesel with C3MR liquefaction. The lowest impact is seen for LBM scenarios with biogas from food waste. The impact of the LBM scenarios is mainly due to carbon monoxide emissions

from combustion in the vehicle’s engine.

The impact on stratospheric ozone depletion is much lower for nearly all LBM scenarios compared to LNG and diesel, especially with system expansion. The variations between different electricity mixes is relatively small, but in many cases the Swedish mix gives the lowest impact. With a German or Italian electricity mix and calculations according to RED, the LBM pathway impact is com-parable to LNG, but still lower than diesel. The processes causing the most ozone layer depletion are the transports and the elec-tricity for AD, upgrading and liquefaction. The LNG scenarios all have a lower impact compared to diesel.

3.2. Economic analysis

Fig. 8andFig. 9show the WTW life cycle cost of LBM scenarios, depending on production capacity, compared to the market price

range of LNG. InFig. 8the LBM production is assumed to be at 95% of

the design production capacity, and inFig. 9the utilization rate is set

to 50% of the capacity. It is clear from thesefigure that LBM

pro-duction has to reach a rather large scale to be able to compete with natural gas without economic incentives. The lower end of the nat-ural gas price range appears to be out of reach for most biomethane technologies, where amine scrubber (AS) and biogas based on food waste is the one that gets closest, although this comparison does not include a sales margin for the LBM. At a production capacity of 120 GWh/year and 95% utilization, this scenario reaches a cost of

0.028 EUR/kWh (Fig. 8). At times when the natural gas price is high,

even WSþ MR with biogas from food waste might be cost-efficient

enough to be competitive, at least at a capacity of 100e120 GWh/ year or more. For the manure scenarios, the production cost at 100e120 GWh/year is still more than twice as high as the LNG price, around 0.050e0.057 EUR/kWh, which means that some form of incentive is necessary for this type of production to be feasible. If the

production capacity is only utilized to 50% (Fig. 9), the specific

in-vestment costs rise and all LBM scenarios end up with production costs higher than the LNG price. For the scenario with AS and food waste, the life cycle cost at 120 GWh/year capacity increases to 0.035 EUR/kWh, while the costs for the manure scenarios reach 0.077e0.084 EUR/kWh at the same plant size.

In the scenarios presented, the LBM is distributed 200 km by semitrailer, and the distribution cost is about 3 EUR/MWh, or 0.3 eurocent per kWh. If the LBM were to be fueled locally rather than being transported from the production site, this cost could be dis-regarded. In that case, the cheapest LBM technology would reach a cost of 0.025 EUR/kWh at 120 GWh/year production capacity and

Fig. 6. Climate change impact of diesel, LNG and LBM well-to-wheel pathways. The LBM pathways are based on biogas from anaerobic digestion of food waste and electricity use is modelled with a German electricity mix.

95% utilization, while WSþ MR with biogas from food waste would nearly match the higher end of the LNG price range (0.030 EUR/ kWh) at 120 GWh/year. If, on the other hand, the distribution range were twice as large (400 km), this would not occur even at

120 GWh/year, and the cost for ASþ MR with biogas from food

waste at 120 GWh/year would go up to 0.031 EUR/kWh.

The rest of the costs for LBM are shared between AD, upgrading

and liquefaction. All specific investment costs decrease with a

Fig. 7. Environmental impact of LBM and LNG pathways compared to diesel in different electricity systems, based on country mix. From top to bottom: Climate change, terrestrial acidification, freshwater eutrophication, ozone formation (human health impact) and stratospheric ozone depletion.

larger scale, particularly for liquefaction and for upgrading by amine scrubbing. Thus, the division of costs between different processes varies with the production capacity. The costs for AD are higher with manure as feedstock than with food waste, as more substrate is treated per kWh biogas produced. With AS the share of costs for AD are higher than with WS, partly because AS requires less electricity and partly because biogas from the AD is assumed to be used for producing process heat for the AS.

4. Discussion

Judging from current trends in policy (Gustafsson and Anderberg,

2020) and use of biomethane in road transports (H€ormann, 2020), it

seems clear that biomethane from second-generation energy sour-ces will have an important place in the work towards cleaner transports. The results of the present study further showdin line

with previous studies by e.g.Shanmugam et al. (2018)andB€orjesson

et al. (2016)dthat liquefied biomethane (LBM) can indeed contribute to reduced environmental impact from heavy-duty vehicles, not least in terms of climate change impact. However, attention must be given to the type of feedstock used and how the electricity used in

biomethane production is generated, which is clear fromFig. 7. With

a lower share of fossil fuels and a lower carbon intensity in electricity production, LBM comes out even better in the comparison with LNG, as the LBM scenarios to a higher degree are dependent on electricity. Conversely, the environmental impact of LBM can be quite high if it is produced with electricity from fossil energy sources. However, the climate change impact is still lower compared to LNG and diesel even in electricity systems with a high share of fossil fuels.

While LBM was found to greatly reduce the climate change impact compared to diesel, the WTW scenarios with LNG resulted in up to 10% higher climate change impact than diesel. This goes

Fig. 8. Life cycle cost of producing LBM as a function of yearly production capacity, compared to market price of LNG, when LBM production is at 95% of the design capacity.

against findings of e.g. Arteconi et al. (2010) (up to 10% lower

climate change impact) and Ou and Zhang (2013)(5e10% lower

impact), but is supported by results from e.g.Stettler et al. (2019)

(up to 8% higher climate change impact) and Cai et al. (2017)

(1e8% higher impact). As pointed out by e.g.Speirs et al. (2020)

andCooper et al. (2019), this has to do with the rate of methane leakage as well as the inferior fuel economy of SI gas engines. Thus, these are two areas of possible technological improvement which can help reduce the environmental impact of vehicles using LNG as well as LBM.

System boundaries and the way in which byproducts are considered can greatly affect the outcome of the calculations in life

cycle assessment (LCA), as also shown by e.g.B€orjesson et al. (2016).

Biomethane produced through AD can be considered to have a positive environmental impact in many impact categories, due to the possibility of displacing mineral fertilizers by utilizing the digestate. However, spreading digestate on agricultural land is not always on option, in which case this type of system expansion is not possible. For example, Dutch regulations do not allow the use of

digestate to replace mineral fertilizer (Pfau et al., 2017). There could

also be other reasons, e.g. economic, for not using the digestate, which would make the RED guidelines more appropriate for calculating the environmental impact. The ISO method, on the other hand, shows what is possible to achieve with a larger system thinking where both the biogas and the digestate from AD are valued and utilized.

Although an LCA can cover many aspects of environmental

impact, it does not fully reflect all the effects of a complex

socio-technical system such as biogas production. Particularly in a local or

regional perspective biogas solutions can give many benefits,

including e.g. waste treatment, increased resource efficiency,

increased yield from agriculture and rural area development (Hagman and Eklund, 2016). A complete analysis of a biogas system would either have to include qualitative elements, for example a multi-criteria analysis, or some way of converting qualitative values into quantitative ones, similar to methods for putting a price on air

pollution, noise and traffic congestions (Bångman and Nordl€of,

2016).

The best-performing liquefaction technology was found to be pressure reduction of methane from the high-pressure gas grid. This is of course provided that the production facility is connected to the gas grid, which is often not the case in some of the Nordic countries (Sweden, Finland, Norway). Furthermore, as shown by

Gustafsson et al. (2020a), adding propane to the biomethane to increase the heating value when injecting it to the gas grid, as has sometimes been custom, greatly increases the environmental impact of this technology and should be avoided. It should also be

taken into account that the share of the gas that can be liquefied

with this technology is limited to around 10e15% (see e.g.He and

Ju, 2013;Tan et al., 2016), leaving most of it to be used in gaseous form. Thus, it would not be the technology of choice if the aim is to

only produce liquefied gas.

In the economic comparison of LBM and LNG, natural gas has the advantage of that it exists in large volumes and requires relatively cheap processes to obtain a high-energy fuel, while biogas is

pro-duced at a much smaller scale with higher speci_{fic costs. Based solely}

on the costs for production and distribution, LBM would not be able to compete with the average market prices of LNG. Considering that the LBM producers would also like to have a certain margin for revenue, it is not realistic that a fair competition would be achievable without economic incentives. The economic calculations by

B€orjesson et al. (2016)suggest that the WTW costs of LBM are more or less in line with market prices of LNG and diesel with the Swedish taxation system. A recent enquiry commissioned by the Swedish government suggested a support package for biogas production that

would amount to up to 0.08 EUR/kWh if the biogas is produced from

manure and is upgraded and liquefied, which is almost double the

existing support (for manure-based biogas) (Westlund, 2019). Such

support systems could turn the competition around and stimulate an increased production of biomethane, although the exact effects on the production in Sweden or other countries would need further investigation.

5. Conclusions

This paper has presented an environmental and economic well-to-wheel analysis of production and distribution pathways of LBM and LNG for heavy-duty trucks, including different feedstock for biogas production, different production technologies and different

electricity mixes. The mainfindings of the study were:

 By replacing diesel with LBM, it is possible to greatly reduce the WTW environmental impact of heavy-duty trucks in terms of

climate change, acidification, eutrophication, ground-level

ozone formation and stratospheric ozone depletion, thus achieving cleaner heavy transports. With a German electricity mix, the climate change impact can be reduced by 45e70% with LBM from manure, and by 50e75% with LBM from food waste, not including the use of digestate. In electricity systems with higher shares of renewable energy, the climate change impact of LBM is even lower.

 If digestate from biogas production is considered to replace mineral fertilizer, in accordance with ISO standards for life cycle

assessment, the climate benefits of using LBM instead of diesel

are even larger. With LBM from manure, the climate change impact could be reduced by 100e125%, and with LBM from food waste by 80e105%.

 Using LNG instead of diesel in heavy-duty trucks does not reduce the WTW climate change impact, but can rather increase

it by up to 10%, due to a lower engine efficiency.

 The environmental impact of LBM is greatly influenced by the type of feedstock used, how the required electricity produced and whether calculations are done according to ISO or RED guidelines.

 Although there are clear environmental incentives to use LBM, it

seems difficult to compete economically with the price of LNG,

due to higher specific production costs.

In order for LBM to compete against LNG and other fossil fuels, some type of economic support is likely to be required. The investigation of how such economic incentives should be formed and what changes might be needed in current practices is however left for future studies. It could also be relevant to look into the

possibility to capture and store or use CO₂ from biogas production,

and what effects that could have on the environmental and eco-nomic performance of biogas systems.

CRediT authorship contribution statement

Marcus Gustafsson: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing

-original draft, Writing - review & editing, Visualization. Niclas

Svensson: Conceptualization, Writing - review& editing,

Visuali-zation, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have

Acknowledgement

This research has received funding from the Swedish Biogas Research Center (BRC), which in turn is funded by the Swedish Energy Agency, grant number 35624-3.

References

Agelbratt, A., Berggren, A., 2015. Potentials of LBM I Europe (No. D3.9). LNG Blue Corridors.

Amon, B., Kryvoruchko, V., Amon, T., Zechmeister-Boltenstern, S., 2006. Methane, nitrous oxide and ammonia emissions during storage and after application of dairy cattle slurry and influence of slurry treatment. Agric. Ecosyst. Environ. 112, 153e162.https://doi.org/10.1016/j.agee.2005.08.030.

Arteconi, A., Brandoni, C., Evangelista, D., Polonara, F., 2010. Life-cycle greenhouse gas analysis of LNG as a heavy vehicle fuel in Europe. Appl. Energy 87, 2005e2013.https://doi.org/10.1016/j.apenergy.2009.11.012.

Bångman, G., Nordl€of, P., 2016. Analysmetod Och Samh€allsekonomiska Kalkylv€ar-den F€or Transportsektorn (No. ASEK 6.0). Trafikverket.

Banks, C.J., Heaven, S., Zhang, Y., Baier, U., 2018. Food waste digestion: anaerobic digestion of food waste for a circular economy (No. 2018:12), 37. IEA Bioenergy Task.

Bauer, F., Hulteberg, C., Persson, T., Tamm, D., 2013. Biogas Upgrading - Review of Commercial Technologies (No. SGC 2013:270). Svenskt Gastekniskt Center. Benjaminsson, J., Nilsson, R., 2009. Distributionsformer f€or biogas och naturgas i

Sverige. Grontmij consulting company.

B€orjesson, P., Lantz, M., Andersson, J., Bj€ornsson, L., M€oller, B.F., Fr€oberg, M., Hanarp, P., Hulteberg, C., Iverfeldt, E., Lundgren, J., R€oj, A., Svensson, H., Zinn, E., 2016. Methane as Vehicle Fuel e a Well to Wheel Analysis (METDRIV) (No. 2016:06). f3 The Swedish Knowledge Centre for Renewable Transportation Fuels.

Cai, H., Burnham, A., Chen, R., Wang, M., 2017. Wells to wheels: environmental implications of natural gas as a transportation fuel. Energy Pol. 109, 565e578. https://doi.org/10.1016/j.enpol.2017.07.041.

Carlsson, M., Uldal, M., 2009. Substrathandbok F€or Biogasproduktion (No. SGC 200). Svenskt Gastekniskt Center.

Cnpc, 2019. A Look at the Development Trend of Liquefaction Technology from China’s LNG Plants.

Cooper, J., Hawkes, A., Balcombe, P., 2019. Life cycle environmental impacts of natural gas drivetrains used in UK road freighting and impacts to UK emission targets. Sci. Total Environ. 674, 482e493. https://doi.org/10.1016/ j.scitotenv.2019.04.091.

Cryo Pur, 2019. Cryo Pur - our solutions [WWW Document]. URL.http://www. cryopur.com/en/our-solutions/#cryofuel(accessed 11.23.19).

Cryostar. Natural gas and small scale liquefaction applications. n.d., [WWW Docu-ment]. URL. https://cryostar.tech/datas-pdf/booklet/en/StarLiteLNG.pdf (accessed 9.12.18).

Daimler, 2018. Natural gas - freightliner trucks [WWW Document]. URL.https:// freightliner.com/trucks/natural-gas/(accessed 9.12.18).

Davidsson, Å., Gruvberger, C., Christensen, T.H., Hansen, T.L., Jansen, J. la C., 2007. Methane yield in source-sorted organic fraction of municipal solid waste. Waste Manag. 27, 406e414.https://doi.org/10.1016/j.wasman.2006.02.013.

De Klein, C., Novoa, R.S.A., Ogle, S., Smith, K.A., Rochette, P., Wirth, T.C., McConkey, B.G., Mosier, A., Rypdal, K., Walsh, M., Williams, S.A., 2006. IPCC Guidelines for National Greenhouse Gas Inventories, Volume 4, Chapter 11: N2O Emissions from Managed Soils, and CO2 Emissions from Lime and Urea Application. IPCC.

Eba, 2020. European Biogas Association Statistical Report: 2019 European Overivew. Brussels, Belgium.

Edwards, R., Larive, J.-F., Rickeard, D., Hass, H., Lonza, L., Maas, H., 2014. Well-to-wheels Report Version 4.a: JEC Well-To-Wheels Analysis : Well-To-Wheels Analysis of Future Automotive Fuels and Powertrains in the European Context. Publications Office of the European Union, Luxembourg.

Eea, 2018. Carbon Dioxide Emissions from Europe’s Heavy-Duty Vehicles. EurObserv’Er, 2017. Biogas Barometer.

European Commission, 2016. Communication from the Commission to the Euro-pean Parliament, the Council, the EuroEuro-pean Economic and Social Committee and the Committee of the Regions - A European Strategy for Low-Emission Mobility.

European Commission, 2009. Directive 2009/28/EC of the European Parliament and of the Council on the promotion of the use of energy from renwable sources. Off. J. Eur. Union 16e62.

Eurostat, 2019a. Shedding light on energy in the EU - a guided tour of energy statistics [WWW Document]. URL. https://ec.europa.eu/eurostat/cache/ infographs/energy/bloc-2c.html.

Eurostat, 2019b. Supply, transformation and consumption of renewables and wastes [nrg_cb_rw] - biogas - Indigenous production [WWW Document]. URL.https:// ec.europa.eu/eurostat/data/database(accessed 7.8.19).

Gasum, 2019. Sweden’s northernmost liquefied gas filling station opens in €Ostersund [WWW Document]. URL.https://www.gasum.com/en/About-gasum/ for-the-media/News/2019/swedens-northernmost-liquefied-gas-filling-station-opens-in-ostersund/(accessed 11.15.19).

Gavrilova, O., Leip, A., Dong, H., MacDonald, J.D., Gomez Bravo, C.A., Amon, B., Barahona Rosales, R., del Prado, A., Aparecida de Lima, M., Oyhantҫabal, W., van der Weerden, T.J., Widiawati, Y., 2006. IPCC Guidelines for National Greenhouse Gas Inventories - Volume 4, Chapter 10: Emissions from Livestock and Manure Management.

Goedkoop, M., Oele, M., Vieira, M., Leijting, J., Ponsioen, T., Meijer, E., 2016. SimaPro Tutorial (No. 5.3). PRe.

Gustafsson, M., Anderberg, S., 2020. Dimensions and characteristics of biogas pol-icies e modelling the European policy landscape. Renew. Sustain. Energy Rev. (accepted for publication).

Gustafsson, M., Cruz, I., Svensson, N., Karlsson, M., 2020a. Scenarios for upgrading and distribution of compressed and liquefied biogas d energy, environmental, and economic analysis. J. Clean. Prod. https://doi.org/10.1016/ j.jclepro.2020.120473.

Gustafsson, M., Svensson, N., Eklund, M., Dahl €Oberg, J., Vehabovic, A., 2020b. Well-to-wheel greenhouse gas emissions of heavy-duty transports depending on electricity carbon intensity. Transport. Res. Transport Environ. (under review). Hagman, L., Eklund, M., 2016. The role of biogas solutions in the circular and

bio-based economy. Biogas Res. Center.

Hagos, D.A., Ahlgren, E.O., 2018. Well-to-wheel assessment of natural gas vehicles and their fuel supply infrastructures e perspectives on gas in transport in Denmark. Transport. Res. Transport Environ. 65, 14e35.https://doi.org/10.1016/ j.trd.2018.07.018.

Hajji, A., Chahartaghi, M., Kahani, M., 2019. Thermodynamic analysis of natural gas liquefaction process with propane pre-cooled mixed refrigerant process (C3MR). Cryogenics 103, 102978. https://doi.org/10.1016/ j.cryogenics.2019.102978.

Hamelin, L., Naroznova, I., Wenzel, H., 2014. Environmental consequences of different carbon alternatives for increased manure-based biogas. Appl. Energy 114, 774e782.https://doi.org/10.1016/j.apenergy.2013.09.033.

He, T.B., Ju, Y.L., 2013. Design and optimization of natural gas liquefaction process by utilizing gas pipeline pressure energy. Appl. Therm. Eng. 57, 1e6. https:// doi.org/10.1016/j.applthermaleng.2013.03.044.

Heisch, P., 2012. Liquefied-to-compressed natural gas - opportunities and strategies. In: Presented at the Gas Show 2012 Exhibition & Conference, Cryostar, Warsawa.

Herbst, K., Mogensen, G., Huber, F., Østberg, M., Skjøth-Rasmussen, M.S., 2010. Challenges in applied oxidation catalysis. Catal. Today 157, 297e302.https:// doi.org/10.1016/j.cattod.2010.01.043.

H€onig, V., Prochazka, P., Obergruber, M., Smutka, L., Kucerova, V., 2019. Economic and technological analysis of commercial LNG production in the EU. Energies 12, 1565.https://doi.org/10.3390/en12081565.

H€ormann, R., 2020. NGVA Europe: Already 17% Renewable Gas Used by the Mobility Sector in Europe.

Igu, 2017. World LNG Report. International Gas Union.

Iso, 2006a. ISO 14040:2006 Environmental Management Life Cycle Assessment -Principles and Framework.

Iso, 2006b. ISO 14044:2006 Environmental Management Life Cycle Assessment -Requirements and Guidelines.

Isuzu, 2018. Isuzu CNG trucks [WWW Document]. URL.https://www.isuzu.com.au/ truck-range/cng-trucks/(accessed 9.12.18).

Iveco, 2015. Iveco eurocargo [WWW Document]. URL.https://www.iveco.com/uk/ products/pages/sustainability-new-eurocargo-2015-iveco.aspx (accessed 9.12.18).

Kampman, B., Leguijt, C., Scholten, T., Tallat-Kelpsaite, J., Brückmann, R., Maroulis, G., Lesschen, J.P., Meesters, K., Sikirica, N., Elbersen, B., 2016. Optimal Use of Biogas from Waste Streams - an Assessment of the Potential of Biogas from Digestion in the EU Beyond 2020. European Commission working group. Karlsson, L.-E., 2018. W€artsil€a.

Kirchmann, H., Witter, E., 1992. Composition of fresh, aerobic and anaerobic farm animal dungs. Bioresour. Technol. 40, 137e142. https://doi.org/10.1016/0960-8524(92)90199-8.

Lantz, M., 2013. Biogas in Sweden - Opportunities and Challenges from a Systems Perspective (Doctoral Thesis). Lund University, Lund, Sweden.

Lantz, M., Bj€ornsson, L., 2016. Emissioner Av V€axthusgaser Vid Produktion Och Anv€andning Av Biogas Från G€odsel (No. 99). Milj€o- och energisystem, LTH, Lunds universitet.

Lantz, M., Ekman, A., B€orjesson, P., 2009. Systemoptimerad Produktion Av For-donsgas - En Milj€o- Och Energisystemanalys Av S€oderåsens Biogasanl€aggning (No. 69). Milj€o- och energisystem, LTH, Lunds universitet.

Larsson, M., Jansson, M., Gr€onkvist, S., Alvfors, P., 2015. Techno-economic assess-ment of anaerobic digestion in a typical Kraft pulp mill to produce biomethane for the road transport sector. J. Clean. Prod. 104, 460e467. https://doi.org/ 10.1016/j.jclepro.2015.05.054.

Ljung, E., Palm, O., Rodhe, L., 2012. €Okad Acceptans F€or Biog€odsel Inom Lantbruket. JTI Swedish Institute for Agricultural and Environmental Technology, Uppsala. Lng Bc, 2013. LNG Blue Corridors. EU FP7, GA No, p. 321592.

Lombardi, L., Francini, G., 2020. Techno-economic and environmental assessment of the main biogas upgrading technologies. Renew. Energy 156, 440e458.https:// doi.org/10.1016/j.renene.2020.04.083.

Lukehurst, C.T., Frost, P., Al Saedi, T., 2010. Utilisation of digestate from biogas plants as biofertiliser. IEA Bioenergy Task 37.

Lyng, K.-A., Brekke, A., 2019. Environmental life cycle assessment of biogas as a fuel for transport compared with alternative fuels. Energies 12, 532.https://doi.org/ 10.3390/en12030532.